This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Herald: The Biostar Herald for Monday, July 22, 2024

The Biostar Herald publishes user submitted links of bioinformatics relevance. It aims to provide a summary of interesting and relevant information you may have missed. You too can submit links here.

This edition of the Herald was brought to you by contribution from Jeremy Leipzig, GenoMax, n.s.j.vandervelden, Rob, and was edited by GenoMax, Istvan Albert,

Just a moment... (community.ukbiobank.ac.uk)

UK Biobank suspends all downloading of data, instead directing all users to migrate to the Research Analysis Platform (UKB-RAP) hosted by DNAnexus.

submitted by: Jeremy Leipzig

Bridge RNAs direct programmable recombination of target and donor DNA | Nature (www.nature.com)

Also: https://www.nature.com/articles/s41586-024-07570-2

Non-CRISPR guided recombination. DNA-based parasite that moves itself around bacterial genomes through a mechanism that hasn't been previously described.

submitted by: GenoMax

x.com (x.com)

After considerably controversy, the publication of a “rebuttal” paper](https://journals.asm.org/doi/10.1128/mbio.01607-23), and a ‘follow-up study, Nature has decided to retract the original article Microbiome analyses of blood and tissues suggest cancer diagnostic approach.

submitted by: Rob

https://academic.oup.com/jamia/advance-article/doi/10.1093/jamia/ocae128/7693164

Comparison of ChatGPT v.3.5 and 4.0 for selected genetics related questions

submitted by: GenoMax

Gene Cluster Visualizations in R • geneviewer (nvelden.github.io)

geneviewer is an R package for plotting gene clusters and transcripts. It imports data from GenBank, FASTA, and GFF files, performs BlastP and MUMmer alignments, and displays results on gene arrow maps. The package offers extensive customization options, including legends, labels, annotations, scales, colors, tooltips, and more.

submitted by: n.s.j.vandervelden

Want to get the Biostar Herald in your email? Who wouldn't? Sign up righ'ere: toggle subscription

herald
• 26,857 views • link
updated by LChart LChart • written by Biostar Biostar

1 answer

Re: Microbiome retraction. I'm really having time squaring the observations in the rebuttal (https://www.nature.com/articles/s41388-024-02974-w) with the ensuing conversation on X.

My reading of the original commentary paper (https://journals.asm.org/doi/10.1128/mbio.01607-23) is that sequences of human origin were not properly filtered, both up-front (through alignment to multiple human references) and competitively (through including human reference(s) in the taxonomy database). This would be the primary flaw. A secondary flaw is also mentioned with regards to normalization of taxonomic data using VOOM, which can impute differences (due to estimated pseudo-library sizes) where none exist.

The rebuttal, in my view, is a tacit agreement that insufficient decontamination was performed -- but that even with (i) extra decontamination (ii) updated normalization + batch correction and (iii) inclusion of human sequences (or removal of contaminated human sequences )in the taxonomy database, there is still a very clear difference in taxa abundance between tumor types (though the specific "driver" taxa do change) -- see "Applying the T2T-KrakenUniq-MicrobialDB pipeline to all of TCGA verifies the existence of cancer type-specific microbiomes"

Yet on X, Salzberg writes:

their response in @oncogenejournal didn't address either of the major flaws we pointed out, but pretended that it did by mis-stating (intentionally, I think) our arguments. Oncogene never asked me for comments

Am I mis-reading the rebuttal? Where is the hoodwink?

One possibility is that they're still simply excluding human sequences from Kraken - but "T2T-KrakenUniq-MicrobialDB" refers to the Salzberg pipeline, which does have these.

I remember some time ago a paper regarding extensive RNA editing being shot down by Pritchard who demonstrated that it was mostly alignment artifacts, and that accounting for the bioinformatic errors rendered the original conclusions completely wrong. That was a no-brainer for retraction -- yet was never retracted

Original https://www.science.org/doi/10.1126/science.1207018

Comment https://www.science.org/doi/10.1126/science.1210484

In this case, it seems like the original conclusions stand, but the specifics of the classifier itself are different. I would be very on-the-fence about issuing a full retraction of the original paper...

• 0 views • link
written by LChart LChart

When I read the piece by Salzberg, it seemed very convincing that the results of the original paper were a consequence of two fundamental mistakes and required both to occur.

I felt that the rebuttal failed to plainly acknowledge either mistake - instead, I argued that the signal is still there and with that, invited the readers to verify themselves.

It took four years to retract this paper, and from my reading between the lines, it seems that Nature themselves did so, and the authors grudgingly went along with it rather than admitting to their errors.

• 0 views • link
written by Istvan Albert Istvan Albert

My problem is there's no way to retract the paper without also retracting the finding.

In this case, the finding (microbial population shifts between tumors) appears to be valid on re-analysis, while the original paper was deeply flawed. Because of this, there should be a mechanism to re-publish corrected analysis. Otherwise anyone interested in the topic will see "retraction" and think the whole idea lacks merit.

• 0 views • link
written by LChart LChart

In my opinion, when scientists publish a finding with a fundamentally wrong analysis, they need to retract the paper.

It is immaterial whether the actual finding happens to be true later and when done correctly.

If the finding is true, then someone else discovering it in a valid manner should get all the credit. It is not realistic to assume that scientists would be discouraged from the field because someone did an analysis in a completely wrong way.

FWIW, whether or not there is such a thing as tumor microbiome seems far from being settled.

Steven Salzberg seems to believe that ALL papers in this field are complete nonsense. Prof. Salzberg is, in my opinion, one of the greatest, if not the greatest, data science-oriented bioinformaticians of all time. So his words carry a lot of weight for me.

and

• 0 views • link
written by Istvan Albert Istvan Albert

Very well articulated.

• 0 views • link
written by LChart LChart

Log in to answer this question.

More posts like this